Unboard train's passengers through doors when the rail line is curved (#14334)

[bcoueraud87]

No description provided.

[namdre]

can you explain why an existing test changed it's behavior? The rail line in that test is not curved and also the door attribute isn't used in the scenario.

[bcoueraud87]

The test I adapted uses doors, look at the input_routes.rou.xml file:

<additional xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="http://sumo.dlr.de/xsd/additional_file.xsd">
    <busStop id="B" lane="a_0" startPos="30" endPos="70">
        <access lane="b_0" pos="48.40"/>
    </busStop>
    <busStop id="C" lane="a_0" startPos="230" endPos="270">
        <access lane="c_0" pos="52.96"/>
    </busStop>
    <busStop id="D" lane="a_0" startPos="430" endPos="470">
        <access lane="d_0" pos="doors"/>
    </busStop>
</additional>

Now, it is true that in this test the rail line is not curved, however, even in the straight case, the computations are now a little bit different.

[namdre]

Sorry, I missed f9b6e68 even though you linked it in #14336 and mistook the test for another one I knew.